Back to EveryPatent.com
| United States Patent |
5,616,876
|
|
Cluts
|
April 1, 1997
|
System and methods for selecting music on the basis of subjective content
Abstract
An interactive network provides music to subscribers. A "more like"
function allows a subscriber to use a seed song to identify other songs
that are similar to the seed song, and to add the new songs to the current
playlist. The similarity between songs is based on the subjective content
of the songs, as reflected in style tables prepared by editors. The
subscriber may control the closeness of the match by adjusting a style
slider provided by the user interface. A style equalizer employs eight
faders that indicate the predominant styles of the songs in the playlist.
A subscriber may use the style equalizer to see what types of songs are
included in the playlist, and to adjust the mix of songs that are played
from the playlist.
| Inventors:
|
Cluts; Jonathan C. (Redmond, WA)
|
| Assignee:
|
Microsoft Corporation (Redmond, WA)
|
| Appl. No.:
|
424781 |
| Filed:
|
April 19, 1995 |
| Current U.S. Class: |
84/609; 84/477R; 434/307A |
| Intern'l Class: |
G09B 015/06; G09B 015/04; G10H 007/00 |
| Field of Search: |
84/609-614,601,602,634-638,477 R,478
358/335
273/433
379/93,96,97,100
434/307 A
|
References Cited
U.S. Patent Documents
| 5250745 | Oct., 1993 | Tsumura | 84/609.
|
| 5454723 | Oct., 1995 | Horii | 84/601.
|
| 5486645 | Jan., 1996 | Suh et al. | 84/610.
|
Other References
The Big Picture, "Introducing Digital Music Express", Georgia Cable TV &
Communications, Apr. 1995.
More Like This, "Get More of What Your're Looking For", Lexis-Nexis, 1995.
|
Primary Examiner: Witkowski; Stanley J.
Attorney, Agent or Firm: Jones & Askew
Claims
What is claimed is:
1. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items from said media server comprising
the steps of:
storing on said server a plurality of programming information items and
editorial data associated with said programming information items;
playing, in response to a first input signal from said input device, an
initial programming information item from said plurality of programming
information items;
creating, in response to a second input signal from said input device, a
list of proposed new programming information items on the basis of said
editorial data associated with said initial programming information item
and said plurality of programming information items;
presenting on said output device said list of said proposed new programming
information items; and
adding, in response to a third input signal from said input device, said
proposed new programming information items to a playlist.
2. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said editorial data comprises subjective content and weighting information
associated with each of said programming information items, and wherein
creating a list of proposed new programming information items comprises
the steps of:
retrieving said editorial data associated with said initial programming
information item;
identifying other programming information items having similar editorial
data; and
selecting a predetermined number of said other programming information
items having similar editorial data.
3. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said editorial data comprises subjective content and weighting information
associated with the author of each of said programming information items,
and wherein creating a list of proposed new programming information items
comprises the steps of:
identifying the author of said initial programming information item;
retrieving the editorial data associated with said author;
identifying other authors having similar editorial data; and
selecting a predetermined number of programming information items by
authors having similar editorial data.
4. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
creating a list of proposed new programming information items comprises
the steps of:
determining the setting of a matching closeness indicator; and
selecting new programming information items by comparing said editorial
data associated with said initial programming information item with said
editorial data associated with said plurality of programming information
items, said comparison being based on the setting of said matching
closeness indicator.
5. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said interactive media distribution system comprises an interactive
television system.
6. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said server is a continuous media server.
7. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said output device is a television monitor and said input device is a
remote control unit.
8. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said programming information items comprise musical selections.
9. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said programming information items comprise movies.
10. In an interactive media distribution system including a media server, a
distribution network, an output device and an input device, a method for
selecting programming information items as recited in claim 1, wherein
said programming information items comprise news stories.
11. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections from said server comprising the steps of:
storing on said server a plurality of musical selections and editorial data
associated with said musical selections;
playing, in response to a first input signal from said input device, an
initial musical selection from said plurality of musical selections;
creating, in response to a second input signal from said input device, a
list of proposed new musical selections on the basis of said editorial
data associated with said initial musical selection and said plurality of
musical selections;
presenting on said output device said list of said proposed new musical
selections; and
adding, in response to a third input signal from said input device, said
proposed new musical selections to a playlist.
12. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections as recited in claim 11, wherein said
editorial data comprises subjective style and weighting information
associated with each of said musical selections, and wherein creating a
list of proposed new musical selections comprises the steps of:
retrieving said editorial data associated with said initial musical
selection;
identifying other musical selections having similar editorial data; and
selecting a predetermined number of said other musical selections having
similar editorial data.
13. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections as recited in claim 11, wherein said
editorial data comprises subjective style and weighting information
associated with an album on which each of said musical selections was
released, and wherein creating a list of proposed new musical selections
comprises the steps of:
identifying the album on which said initial musical selection was released:
retrieving said editorial data associated with said album;
identifying other albums having similar editorial data; and
selecting a predetermined number of musical selections from said other
albums having similar editorial data.
14. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections as recited in claim 11, wherein said
editorial data comprises subjective style and weighting information
associated with the artist who performed each of said musical selections,
and wherein creating a list of proposed new musical selections comprises
the steps of:
identifying the artist who performed said initial musical selections;
retrieving the editorial data associated with said artist;
identifying other artists having similar editorial data; and
selecting a predetermined number of musical selections performed by artists
having similar editorial data.
15. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections as recited in claim 11, wherein creating a
list of proposed new musical selections comprises the steps of:
determining the setting of a matching closeness indicator; and
selecting new musical selections by comparing said editorial data
associated with said initial musical selection with said editorial data
associated with said plurality of musical selections, said comparison
being based on the setting of said matching closeness indicator.
16. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections as recited in claim 11, wherein said
interactive music distribution system comprises an interactive television
system.
17. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections as recited in claim 11, wherein said server
is a continuous media server.
18. In an interactive music distribution system including a server, a
distribution network, an output device and an input device, a method for
selecting musical selections as recited in claim 11, wherein said output
device is a television monitor and said input device is a remote control
unit.
19. A method for classifying and selecting programming information items
having subjective content, comprising the steps of:
storing a plurality of programming information items;
storing editorial data associated with said programming information items,
said editorial data including a plurality of categories and weightings
associating each programming information item with said categories;
identifying, in response to a first input signal from an input device, an
initial programming information item from said plurality of programming
information items;
determining, in response to a second input signal from said input device,
the setting of a matching closeness indicator;
determining matching categories for said initial programming entry, said
matching categories including the categories whose weightings correspond
to the position of the matching closeness indicator;
identifying matching programming information items, said matching
programming information items including said matching categories with
weightings corresponding to the setting of said matching closeness
indicator; and
presenting said matching programming information items of a user.
20. A method for classifying and selecting programming information items
having subjective content as recited in claim 19, wherein said matching
closeness indicator is set in response to a third input signal from said
input device.
21. A method for classifying and selecting programming information items
having subjective content as recited in claim 19, wherein said programming
information items comprise songs.
22. A method for classifying and selecting programming information items
having subjective content as recited in claim 19, wherein said programming
information items comprise news stories.
23. A method for classifying and selecting programming information items
having subjective content as recited in claim 19, wherein said programming
information items comprise movies.
24. A system for classifying and selecting programming information having
subjective content, comprising:
a data storage device containing a plurality of programming information
items and editorial data associated with said programming information
items;
an output device for providing information to a user;
an input device for receiving input from said user; and
a computer associated with said data storage device, said computer being
configured to:
play, in response to a first input signal from said input device, an
initial programming information item from said plurality of programming
information items,
create, in response to a second input signal from said input device, a list
of proposed new programming information items on the basis of said
editorial data associated with said programming information items,
present on said output device said list of proposed new programming
information items, and
add, in response to a third input signal from said input device, said
proposed new programming information items to a playlist.
25. A system for classifying and selecting programming information as
recited in claim 24, wherein said editorial data includes a plurality of
style categories and weightings associated with each of said style
categories.
26. A system for classifying and selecting programming information as
recited in claim 25, wherein said proposed new programming information
items and said initial programming selection include at least one
identical style category.
27. A system for classifying and selecting programming information as
recited in claim 24, wherein said programming information items comprise
songs.
28. A system for classifying and selecting programming information as
recited in claim 24, wherein said programming information items comprise
new stories.
29. A system for classifying and selecting programming information as
recited in claim 24, wherein said programming information items comprise
movies.
30. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist, comprising the steps of:
loading a playlist including a plurality of programming information items;
loading editorial data associated with said plurality of programming
information items;
displaying on said output device a predetermined number of indicators;
associating with each of said indicators a category from said editorial
data, said indicators being positioned to indicate the portion of said
plurality of programming information items corresponding to each of said
categories;
adjusting, in response to an input signal from said input device, the
position of at least one of said indicators;
selecting programming information items from said playlist such that the
portions of said selected programming information items associated with
each of said categories corresponds to the adjusted positions of said
indicators; and
playing said selected programming information items on said output device.
31. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, wherein associating a
category with each of said indicators comprises the steps of:
determining the predominant categories of said categories; and
assigning said predominant categories to said indicators.
32. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 31, wherein determining the
predominant categories comprises the steps of:
identifying the categories included in said editorial data; and
determining the number of programming information items associated with
each of said categories.
33. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, wherein selecting songs
comprises the steps of:
determining a total number of positions associated with said indicators;
determining a number of positions associated with one of said indicators;
and
dividing said number of positions by said total number of positions.
34. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, where;in said
programming information items comprise songs.
35. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, wherein said programming
information items comprise news stories.
36. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, wherein said programming
information items comprise movies.
37. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, wherein said server
comprises a desktop computer and said output device comprises a display.
38. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, wherein said input
device is a remote control unit.
39. In an interactive system including a server, an output device, and an
input device, a method for indicating the mix of programming information
included in a playlist and adjusting the mix of programming information
played from said playlist as recited in claim 30, wherein said video
output device is a television monitor.
40. In an interactive music system including a server, a video output
device, an audio output device and an input device, a method for
indicating the mix of songs included in a playlist and adjusting the mix
of songs played from said playlist, comprising the steps of:
loading a playlist including a plurality of songs;
loading editorial data associated with said plurality of songs;
displaying on said video output device a predetermined number of
indicators;
associating with each of said indicators a category from said editorial
data;
setting the position of said indicators to indicate the portion of said
plurality of songs corresponding to each of said categories;
adjusting, in response to an input signal from said input device, the
position of at least one of said indicators;
selecting songs from said playlist such that the portions of said selected
songs associated with each of said categories corresponds to the adjusted
positions of said indicators; and
playing said selected songs on said audio output device.
41. In an interactive music system including a server, a video output
device, an audio output device and an input device, a method for
indicating the mix of songs included in a playlist and adjusting the mix
of songs played from said playlist as recited in claim 40, wherein
associating a category with each of said indicators comprises the steps
of:
determining the predominant categories of said categories; and
assigning said predominant categories to said indicators.
42. In an interactive music system including a server, a video output
device, an audio output device and an input device, a method for
indicating the mix of songs included in a playlist and adjusting the mix
of songs played from said playlist as recited in claim 41, wherein
determining the predominant categories comprises the steps of:
identifying the categories included in said editorial data; and
determining the number of songs associated with each of said categories.
43. In an interactive music system including a server, a video output
device, an audio output device and an input device, a method for
indicating the mix of songs included in a playlist and adjusting the mix
of songs played from said playlist as recited in claim 40, wherein
selecting songs comprises the steps of:
determining a total number of positions associated with said indicators;
determining a number of positions associated with one of said indicators:
and
dividing said number of positions by said total number of positions.
44. In an interactive music system including a server, a video output
device, an audio output device and an input device, a method for
indicating the mix of songs included in a playlist and adjusting the mix
of songs played from said playlist as recited in claim 40, wherein said
input device is a remote control unit.
45. In an interactive music system including a server, a video output
device, an audio output device and an input device, a method for
indicating the mix of songs included in a playlist and adjusting the mix
of songs played from said playlist as recited in claim 40, wherein said
video output device is a television monitor.
Description
TECHNICAL FIELD
The present invention relates to systems and methods for selecting and
playing audio selections, and more particularly relates to methods for
selecting and playing audio selections on the basis of their subjective
content.
BACKGROUND OF THE INVENTION
The expansion and improvement of cable television systems (sometimes
referred to as community antenna television or CATV systems) have made it
possible for cable companies to provide a variety of programming services
to subscribers. These services typically include a multitude of television
channels that are viewed on the subscriber's television. Some cable
companies also provide music channels that are connected to a subscriber's
stereo system through a subscriber terminal.
Although CATV systems were originally designed to distribute television
signals in the "downstream" direction only (i.e., from a central "headend"
location to multiple subscriber locations, which is also known as the
"forward" path), the advent of pay-per-view services and of other
interactive television applications has fueled the development of
bidirectional or "two-way" cable systems. These two-way cable systems also
provide for the transmission of signals from the subscriber locations back
to the headend via an "upstream" direction or a "reverse" path.
By upgrading conventional CATV systems to increase their bandwidth, cable
service providers can use the additional channels gained by this wider
bandwidth network to provide many new subscriber services. The
ever-expanding deployment of fiber optic technology supports the
implementation of an "interactive network" that allows a subscriber to
obtain desirable services or programming at a time and date specified by
the subscriber. Indeed, it is feasible that this interactive network will
have sufficient bandwidth to supply hundreds of channels of programming
information, thereby leading to an explosion of program options available
to subscribers. Potential subscriber services supported by this
interactive network include Movies on Demand (MOD) or Video on Demand
(VOD), interactive music channels, interactive computing, shopping,
entertainment, and other related services.
An interactive network makes it possible for subscribers to have immediate
access to vast selections of music. For example, record companies may
provide catalogs of their music for subscribers to listen to via an
interactive network. Similarly, various publishers may compile playlists
of various styles of music (e.g., Jazz, Classical, Top 40, etc.) that will
be available to subscribers via an interactive network.
When music catalogs are available via an interactive network, a subscriber
must have a way to select the music he or she would like to listen to. The
computers that form a part of an interactive network facilitate selection
by song title, artist, or album name. As in a record store, music may also
be classified and searched by style (e.g., Jazz, Classical, Top 40, etc.).
Thus, in an interactive network, it will be very simple for a subscriber
to select a specific song.
However, unless a subscriber is familiar with a particular artist or song
title, there is no simple way to identify other music that the subscriber
may enjoy. Because of the subjective nature of such a decision, there is
no simple way for a subscriber to identify additional music that is
similar to a song he or she likes. This is a significant disadvantage in
an environment where a large assortment of music is readily available.
When a listener browses published playlists, the playlists are typically
described by a short title, such as Jazz, Classical, Top 40, Progressive
Rock, etc. When a subscriber listens to such a playlist, there is no
simple way for the subscriber to get a clearer idea of the specific types
of music that are included in the playlist. Similarly, there is no simple
way for a subscriber to alter the mix of the songs that are played back
from the playlist.
In summary, there is no simple, effective way for a subscriber to identify
and select music he or she is likely to enjoy on the basis of the music's
subjective content and its similarity to a song the subscriber is familiar
with. Furthermore, there is no way for a user to quickly assess the mix of
music included in a playlist and to alter the mix of music played from the
playlist.
Therefore, there is a need in the art for a system that allows a subscriber
to pick a song he or she likes and to then identify additional songs that
include similar subjective content. Likewise, there is a need in the art
for a system that allows a user to perceive the content of a playlist and
alter the mix of songs played from the playlist.
SUMMARY OF THE INVENTION
The present invention satisfies the above described needs by providing
systems and methods for selecting and playing music based on its
subjective content.
Generally described, the present invention provides a method for selecting
programming information items in an interactive media distribution system
that includes a server, a distribution network, an output device and an
input device. The method includes storing on the server a plurality of
programming information items and editorial data associated with the
programming information items. An initial programming information items is
played in response to a first input signal. In response to a second input
signal, a list of proposed new programming information items is created on
the basis of the editorial data associated with the initial programming
information item and the plurality of programming information items. The
list of proposed new programming information items is presented on the
output device. The proposed new programming information items are then
added to a playlist in response to a third input signal.
The present invention also provides a method for classifying and selecting
programming information items having subjective content. A plurality of
programming information items and editorial data associated with the
programming information items are stored. The editorial data includes a
plurality of categories and weightings associating each programming
information item with the categories. An initial programming information
item is selected in response to a first input signal. The setting of a
matching closeness indicator is determined in response to a second input
signal. The method determines matching categories for the initial
programming entry. The matching categories include the categories whose
weightings correspond to the position of the matching closeness indicator.
The method determines matching programming information items based on the
initial programming information item. The matching items include the
matching categories with weightings corresponding to the setting of the
matching closeness indicator. The matching items are presented to the
user.
The present invention also provides a system for classifying and selecting
programming information having subjective content. The system includes a
data storage device containing a plurality of programming information
items and editorial data associated with the programming information
items, an output device for providing information to a user, an input
device for receiving input from the user, and a computer associated with
the data storage device. The computer is configured to play an initial
programming information item in response to a first input signal. The
computer creates a list of proposed new programming information items on
the basis of the editorial data associated with the programming
information items in response to a second input signal. The list of
proposed new programming information items is presented on the output
device. Finally, the proposed new programming information items are added
to a playlist in response to a third input signal.
In another aspect, the present invention provides a method for indicating
the mix of programming information included in a playlist and adjusting
the mix of programming information played from the playlist. The method
includes loading a playlist including a plurality of programming
information items and loading editorial data associated with the
programming information items. A predetermined number of indicators are
displayed on the output device. Each of the indicators is associated with
a category from the editorial data. The indicators are positioned to
indicate the portion of the plurality of programming information items
corresponding to each of the categories. At least one of the indicators is
adjusted in response to an input signal from an input device. In response
to the adjustment of one or more indicators, the method selects
programming information items from the playlist such that the portions of
the selected programming information items associated with each of the
categories corresponds to the adjusted positions of the indicators.
It is therefore an object of the present invention to provide a system for
classifying and selecting information having subjective content.
It is another object of the present invention to provide a method for the
context based selection of subjective material.
It is another object of the present invention to provide a method for
adding items having subjective content to a group of items having similar
subjective content.
It is another object of the present invention to predict, based on a
listener's current choice of music, the choices from an audio content
database that are most like the current choice.
It is another object of the present invention to identify other music that
is similar to the music a listener is listening to.
It is another object of the present invention to identify more music that
is like a current musical selection.
It is another object of the present invention to identify other movies that
are similar to a movie a viewer is watching.
It is another object of the present invention to display the types of music
are in a playlist.
It is another object of the present invention to allow a user to alter the
mix of music that is selected and played from a playlist.
It is another object of the present invention to select various types of
programming on the basis of its subjective content.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 illustrates an interactive network system.
FIG. 2 illustrates a set top terminal, which forms a part of the
interactive network system of FIG. 1.
FIG. 3 illustrates the preferred remote control unit for use with the set
top terminal of FIG. 2.
FIG. 4 is illustrates the features of the initial screen display in the
preferred audio on demand system.
FIG. 5 illustrates the features of the playlist screen display in the
preferred audio on demand system.
FIG. 6 illustrates a list of songs provided in response to the find button
on the playlist screen display of FIG. 5.
FIG. 7 illustrates the "more like" panel provided in response to the "more"
button on the playlist screen display of FIG. 5.
FIG. 8 illustrates a list of songs provided by the "more like" function.
FIG. 9 is a flow diagram illustrating the steps taken by a subscriber when
using the "more like" function.
FIG. 10 is a flow diagram illustrating the "more like" function as
implemented in a program module running on the preferred interactive
network.
FIG. 11 illustrates the features of the "style equalizer" screen display.
FIG. 12 is a flow diagram illustrating the steps taken by a subscriber when
using the "style equalizer" function.
FIG. 13 is a flow diagram illustrating the "style equalizer" function as
implemented in a program module running on the preferred interactive
network.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
The preferred embodiment of the present invention is directed to systems
and methods for selecting music on the basis of its subjective content,
and is implemented in an interactive network system that can deliver a
variety of services, including entertainment, information, and transaction
services, to consumers via an interactive broadband network. The preferred
system, which is referred to as the audio on demand system, allows a
subscriber to listen to songs provided by the system. The subscriber may
select songs on the basis of title, artist and album. The subscriber may
also select playlists, which arc predetermined collections of songs The
audio on demand system provides a "more like" function that identifies
more music that is like the subscriber's current selection. In addition,
the system includes a "style equalizer" that allows a subscriber to see
the predominant styles of music included in a playlist, and to adjust the
mix of music played from the playlist.
Although the preferred embodiment will be generally described in the
context of an interactive television system for delivering broadcast
television programs, music, and related information, those skilled in the
art will recognize that the present invention also can be used to support
the delivery of other forms of programming information, including radio,
broadcast print, audio, games, computer software, including program
modules such as application programs and operating systems, and other
combinations of audio, video and/or computer software. Accordingly, it
will be understood that the terms "programming information" and
"programming information items" generally include information transmitted
electronically to entertain, instruct, educate, or inform the recipient,
as well as program modules for supporting these services.
Turning first to the nomenclature of the specification, the detailed
description which follows is represented largely in terms of processes and
symbolic representations of operations by conventional computer
components, including a central processing unit (CPU), memory storage
devices for the CPU, and connected pixel-oriented display devices. These
operations include the manipulation of data bits by the CPU and the
maintenance of these bits within data structures resident in one or more
of the memory storage devices. Such data structures impose a physical
organization upon the collection of data bits stored within computer
memory and represent specific electrical or magnetic elements. These
symbolic representations are the means used by those skilled in the art of
computer programming and computer construction to most effectively convey
teachings and discoveries to others skilled in the art.
For the purposes of this discussion, a process is generally conceived to be
a sequence of computer-executed steps leading to a desired result. These
steps generally require physical manipulations of physical quantities.
Usually, though not necessarily, these quantities take the form of
electrical, magnetic, or optical signals capable of being stored,
transferred, combined, compared, or otherwise manipulated. It is
conventional for those skilled in the art to refer to these signals as
bits, values, elements, symbols, characters, terms, numbers, records,
files or the like. It should be kept in mind, however, that these and
similar terms should be associated with appropriate physical quantities
for computer operations, and that these terms are merely conventional
labels applied to physical quantities that exist within and during
operation of the computer.
It should also be understood that manipulations within the computer are
often referred to in terms such as adding, comparing, moving, etc. which
are often associated with manual operations performed by a human operator.
It must be understood that no involvement of a human operator is necessary
or even desirable in the present invention. The operations described
herein are machine operations performed in conjunction with a human
operator or user that interacts with the computer. The machines used for
performing the operation of the present invention include general purpose
digital computers or other similar computing devices.
In addition, it should be understood that the programs, processes, methods,
etc. described herein are not related or limited to any particular
computer or apparatus. Rather, various types of general purpose machines
may be used with programs constructed in accordance with the teachings
described herein. Similarly, it may prove advantageous to construct
specialized apparatus to perform the method steps described herein by way
of dedicated computer systems with hard-wired logic or programs stored in
nonvolatile memory, such as read only memory.
Referring now the drawings, in which like numerals represent like elements
throughout the several figures, the present invention and the preferred
operating environments will be described.
THE OPERATING ENVIRONMENT
A typical CATV system for the delivery of television programming to
subscribers comprises three main elements: a headend, a distribution
system, and subscriber drops.
The "headend" is a signal reception and processing center that collects,
organizes and distributes signals. The headend receives
satellite-delivered video and audio programming, over-the-air broadcast
television station signals, and network feeds delivered by terrestrial
microwave and other communication systems. In addition, headends may
inject local broadcast programming into the package of signals sent to
subscribers, such as commercials and live programs created in a television
studio.
The "distribution system" carries the signals from the headend to a number
of distribution points in a community and, in turn, distributes the these
signals to individual neighborhoods for delivery to subscribers. A modern
distribution system typically comprises a combination of coaxial cable and
optical fibers with trunk amplifiers periodically spaced to compensate for
attenuation of the signals along the line.
"Subscriber drops" are taps in the distribution system that feed individual
lines into subscribers' television sets or subscriber set-top terminals,
often referred to as "subscriber premises equipment" or "customer premises
equipment" ("CPE").
Referring to FIG. 1, an interactive network system 10 includes a headend
system 12 for delivering programming information to and receiving
instructions from a consumer system 14 via a "two-way" distribution
network 16. The headend system 12 is the control center for collecting,
organizing, and distributing the signals for all interactive network
operations and the source for all programming information. The
distribution network 16 transports signals carrying programming
information and instructions between the headend system 12 and the
consumer system 14. The distribution network 16 can include a world-wide
public asynchronous transfer mode (ATM) compatible network with links to
the Internet, third party service providers, and other wired and wireless
communications networks. The consumer system 14 includes the equipment
required for a consumer to receive programming information directly at his
or her office or residence and to transmit requests and instructions to
the headend system 12.
The headend system 12 can include a set of headend servers 20, including a
continuous media server (CMS) system 22 and one or more administrative
servers 24, to support various network functions, and a control network 26
linking these headend servers. The headend servers 20 can execute program
modules, including service and application program software, to support
the transmission of programming information and the reception of requests
for such programming information.
It will be appreciated that the headend servers 20 are not necessarily
located in one physical location, but can be linked by wired and/or
wireless communications paths supplied by the control network. The control
network 26 can be a local area network, a wide area network, or a
combination of both types of networks. For the preferred embodiment, the
control network 26 is implemented as an ATM-based network for routing
digital data between the headend servers 20 and the distribution network
16.
The CMS system 22 is a server-based file storage and delivery system that
can manage on-demand access to stored digitized data, such as audio and
video. On-demand access of digitized data is a particularly desirable
characteristic of the CMS system 22 because it allows the interactive
network to support the on-demand delivery of various types of programming,
such as music, movies, etc. The preferred CMS system 22 can supply digital
data streams at a constant rate to numerous consumers of the consumer
system 14.
The CMS system 22 includes one or more storage servers 28, which operate to
retrieve and to transmit the digitized data as required by clients of the
CMS system, i.e., the equipment of the consumer system 14. The digitized
data, which typically comprises programming information, is maintained on
one or more memory storage devices 30 connected to the storage servers 28.
Each memory storage device 30 can be implemented as a SCSI hard disk
drive, an optical storage system, or any other similar mass storage media.
By spreading the data management operations across a group of storage
servers and memory storage devices, user load can be balanced with the
limited disk, network, and input/output (I/O) resources of the headend
system. This also supports fault tolerance by replicating digitized data
within the CMS system 22 to survive the failure of a storage server or a
memory storage device.
To support the tasks of updating or revising programming information stored
on a memory storage device 30 of the CMS system 22, a computer workstation
32 and a remote server 34 can be connected to the control network 26 via a
communications link 36. This communications link allows a program
distributor or supplier, which typically operates at a location remote
from the CMS system 22, to transmit programming information for storage by
one or more of the memory storage devices 30 and eventual distribution to
consumers via the headend system 12. The communications link 36 can be
implemented by either a wireless or wired communications system. For
example, the communications link 36 can be constructed as a microwave link
or as a conventional telephone link.
The administrative servers 24 of the headend system 12 can support a
variety of services and applications associated with the interactive
network system 10, including network security, monitoring, object storage,
financial transactions, data management, and other administrative
functions. The administrative servers 24 also handle the interactive
service requests or instructions transmitted via the consumer system 14 by
consumers. For an application involving a large base of consumers, an
administrative server 24 is preferably dedicated to a particular service
or function. For example, one or more servers can handle all consumer
authorization requirements, whereas other servers can handle network
management services, and so forth. These administrative servers preferably
support the Simple Network Management Protocol (SNMP) to enable end-to-end
network administration and monitoring.
The headend system 12 also can support the distribution of programming
information and other services via an analog distribution system 38 that
is coupled to the distribution network 16. This distribution of analog
formatted signals can be handled by a separate headend system associated
with a community antenna television (CATV) system. The headend of the CATV
system typically supports satellite-delivered video and audio programs,
over-the-air broadcast television station signals, and broadcast network
signal feeds delivered by microwave and other communications systems.
The distribution network 16 is a two-way communications network that
connects the headend system 12 to various community distribution points of
the consumer system 14 and, in turn, to individual neighborhood nodes for
delivery to consumers of services supplied by the interactive network
system 10. The distribution network 16 comprises one or more downstream
channels supporting transmissions from the headend system to the consumer
system and one or more upstream channels for carrying transmissions from
the consumer system to the headend system. This bidirectional
communications network supports delivery of programming information via
the headend system 12 to each consumer and the delivery of requests for
programming information by a consumer to the headend system 12. The
distribution network 16 can be implemented by a microwave distribution
system, a telephone system, coaxial cables, optical fibers, or any
combination of these delivery systems. However, the preferred distribution
network is implemented by a combination of hybrid optical fiber/coaxial
cable (HFC) and optical fiber-to-the-curb (FTTC).
Those persons skilled in the art will appreciate that the programming
information delivered over the distribution network 16 typically includes
both video and audio signals. Programming information can be delivered in
digital format, analog format, or a combination of both analog and digital
formats. For the preferred embodiment, music-related programming is
delivered as a stream of digital audio and video signals in a compressed
digital data stream, which may include conventional MPEG-1 and MPEG-2
compressed video streams. Likewise, requests or instructions issued by
consumers via the consumer system 14 are preferably formatted as digital
signals.
The CMS system 22 and the administrative servers 24 are connected to the
distribution network 16 via an ATM switching system 40. The ATM switching
system 40 supports network switching requirements for delivery by the
headend system 12 of digital data streams carrying multimedia content and
the handling of interactive service requests from consumers.
Because the interactive network 10 is a two-way communications system, the
ATM switching system 40 preferably connects to the distribution network 16
via modulation/demodulation devices. The downstream channels of the
distribution network 16 can be connected to the ATM switching system 40
via digital modulators 42, whereas the reverse channels of the
distribution network 16 are connected to reverse channel receivers 44.
Each consumer within a neighborhood node of the consumer system 14 is
connected to the distribution network 16 via a subscriber drop cable 46,
which is typically part of a local cable network administered by a
multiple service operator (MSO). The drop cable 46 is typically a coaxial
cable or optical fiber connected to a set-top terminal 48 or set-top box
located at the consumer's location. This combination of the drop cable 46
and the set-top terminal 48 operates as a "tap" into the distribution
network 16, and allows the consumer to (1) receive program modules and
programming information distributed by the headend system 12 and to (2)
transmit requests or instructions to the headend system 12. For example,
the set-top terminal 48 can accept and convert signals carrying
programming information to a format compatible for presentation by an
output device 50, such as a television or a computer system. This output
device 50, which can connected to the set-top terminal via a conductive
path 52 such as coaxial cable, preferably includes a receiver and a
display or monitor for receiving and displaying programs and
program-related information. Those skilled in the art will understand that
the output device 50 can be implemented as a combination of separate
components, such as a receiver and a monitor, or as a single component,
such as a conventional television or a general purpose computer system.
Selected operating functions of the set-top terminal 48 can be controlled
by an input device 54 capable of supplying input data to the set-top
terminal 48. The input device 54 can be used to transmit command signals
to the set-top terminal 48 and to input character-based data, such as
text, for processing by the set-top terminal 48. For example, the input
device 54 can be used to control the position of a display object
presented by the output device or to enter text for conducting a
service-related transaction supported by the interactive network 10. The
input device 54 can be implemented as one or more devices for inputting
data, including a hand held control, a keyboard, a mouse device, a game
control, a joystick, a pen or stylus, a trackball, or a track pad.
For the preferred embodiment, the input device 54 is implemented as a hand
held remote control unit capable of transmitting infrared signals carrying
commands for controlling the operation of the set-top terminal 48. The
remote control unit can include a directional keypad having distinct keys
for allowing the user to control direction (up, down, left, right) and
relative changes in volume or channel (increase or decrease), as well as
absolute changes to channel value via a numeric key pad. The remote
control unit and its functions are more fully described in conjunction
with FIG. 3.
FIG. 2 illustrates the basic components of the set-top terminal 48. The
primary components of the set-top terminal 48 include a network receiver
56, a network transmitter 58, a remote control receiver 60, a central
processing unit (CPU) 62, and memory 64. These components are connected by
a system bus 70, which can carry control, address, and data signals. The
network receiver 56 conducts tuning operations for receiving a selected
channel of the interactive network 10 and decoding operations for decoding
compressed digitized data supplied via the interactive network 10. For
example, the set-top terminal 48 can include MPEG decoding capability for
converting the compressed digitized data into standard National Television
Standard Committee (NTSC) video signals for reception by a conventional
television. The network transmitter 58 transmits requests for programming
information and related instructions for processing by the headend system
12. The network receiver 56 and the network transmitter 58 can be
connected to the distribution network 16 via the drop cable 46. The remote
control receiver 60, which is preferably implemented as an infrared
receiving device, can decode signals carrying the commands issued by the
input device 54, such as a remote control unit 80.
The CPU 62, which is connected to the network receiver and transmitter 56
and 58, as well as to the remote control receiver 60, controls the
operations of the set-top terminal 48 and supports the rendering of
graphical images that form a part of the user interface. The CPU 62 is
typically implemented by at least one microprocessor, such as the model
80486 or the "PENTIUM" microprocessor, manufactured by Intel Corporation,
Santa Clara, Calif. The CPU 62 communicates, by means of control, address,
and data signals, with the remaining components of the set-top terminal 48
through the system bus 70. The CPU 62 operates in conjunction with the
operating system 66 to retrieve, process, store, and display data. It will
be appreciated that the processing functions of the CPU 62 may be divided
among two or more microprocessors to support the presentation of a
graphics-intensive user interface. For example, a microprocessor may be
dedicated to control operations associated with the bi-directional
communications with the headend system 12, whereas another microprocessor
may be dedicated to the generation of graphics.
The memory 64, which is connected to the CPU 62, is useful for storing one
or more program modules and data associated with set-top terminal
operations. Program modules stored in the memory 64 can include operating
system 66 and one or more application programs 68. The memory 64 can be
implemented as a combination of dynamic memory, such as random access
memory (RAM), and static memory, such as read only memory (ROM).
The operating system 66 comprises a set of computer programs that control
the internal functions of the set-top terminal and support the execution
of other program modules, including application programs 68. The preferred
operating system 66 supports a graphics-based presentation of
program-related information, including control items that visually
represent control functions of the operating system and other program
modules. A control item or control object is any visual image that can be
manipulated by the user to perform an operation. The operating system 66
can receive and interpret input data supplied by the input device 54, as
received by the remote control receiver 60. As described in more detail
below, a user can "select" and "activate" (or launch) control items by the
use of the input device 54 in a manner similar to the computer arts.
For the preferred set-top terminal 48, the memory includes a ROM containing
at least a portion of program module representing "boot code" 72 for
initializing the operations of the set-top terminal 48. Upon power-up of
the set-top terminal 48, the boot code 72 initiates a request for the
headend system 12 to download certain program modules, including the
operating system 66 and one or more application programs 68. The program
modules can be stored within the memory 64 of the set-top terminal 48.
This downloading process allows the headend system 12 to easily update the
program modules used in set-top terminals 48 throughout the interactive
network 10. For example, the application programs 68 may be maintained
within the set-top terminal 48 only during actual use of the features of
these programs; otherwise, these application programs are maintained at
the headend system 12. Thus, it will be appreciated that the preferred
set-top terminal 48 relies heavily upon data storage mechanisms located at
the headend system 12 rather than within the set-top terminal 48 itself.
The set-top terminal 48 can be connected to a peripheral device via
input/output (I/O) ports 74. The I/O ports 74 support the connection of
the system bus 70 to a connected peripheral device. For example, the
output device 50 can be connected to the I/O ports 74 via a conductor 52.
Likewise, an input device 54, such as a game control 90, can be connected
to the I/O ports 74. In contrast to the remote control unit 80, which
communicates with the remote control receiver 60 via a wireless
communications link, other types of input devices 54 are typically
connected to the I/O ports 74 via a cable. Nevertheless, those skilled in
the art will appreciate that input devices 54 can communicate with the
set-top terminal 48 by use of either wireless or wired communications
links.
Generally, when a user first powers-up a set-top terminal 48, the set-top
terminal 48 contacts the headend system 12 and requests the downloading of
certain program modules, including the operating system 66. In response to
loading these program modules, the set-top terminal 48 enters a stand-by
mode to limit power consumption and awaits a command signal initiated by a
user pressing a key or button on an input device 54, such as a remote
control unit 80. In this stand-by mode, the set-top terminal can
communicate with the headend system and can respond to administrative
requests transmitted by the headend system 12. In the event that a user
tunes to an interactive channel (such as the audio on demand service), the
set-top terminal 48 changes modes and enters the active mode. In the
active mode, the set-top terminal 48 communicates with the headend system
12 to process the instructions transmitted by the remote control unit. For
example, the set-top terminal 48 responds to a command requesting
programming information by forwarding this instruction to the headend
system 12 via the drop cable 46 and the distribution network 16. The
headend system 12 responds by retrieving selected programming information
from the CMS system 22 and transmitting the selected programming
information via the return path provided by the distribution network 16
and the drop cable 46. The set-top terminal then supplies this programming
information in the proper format for presentation by the display 50.
FIG. 3 illustrates the preferred remote control unit 80, which is used to
transmit commands to the set-top terminal 48. The remote control unit 80
includes a variety of keys that are common to remote control units for use
with conventional television sets. These include power on/off 100, channel
up 102, channel down 104, volume up 106, volume down 108, mute 110, and a
10 digit numeric keypad 112.
The preferred remote control unit also includes keys that are specifically
related to preferred interactive system. A menu button 114 is used to open
and close on-screen menus. A directional control 116 is a rocker switch
that is used to select specific control items by moving a cursor up, down,
left or right. An action button 118 is used to activate a selected control
item. A help key 120 is used to initiate on-screen help. An "A" button 122
and "B" button 124 are used to select specific options that are provided
in some contexts.
The Preferred Systems and Methods for Selecting Music Based on Subject
Content
Turning now to FIGS. 4-13, the preferred systems and methods for selecting
music based on subjective content will be described. The primary features
of the present invention are a "more like" music search function and a
"style equalizer" (style EQ). In addition, the preferred audio on demand
system also provides a variety of ancillary features. These features allow
a subscriber to find a specific song by artist and title or select a
playlist, which is a collection of songs. Once a playlist is selected, the
user may review the contents of the playlist and select another song in
the playlist. While a song is playing, the listener may add the song to a
playlist called "my favorites", or mark the song so that it is never
played again. Each of these features is discussed in conjunction with its
corresponding user interface and control objects.
FIGS. 4-6 illustrate general aspects of the user interface employed in the
preferred embodiment of the present invention, which is referred to as the
audio on demand system. FIGS. 7-10 include screen displays and flow charts
associated with the "more like" function, which provides the subscriber
with more songs that are like the current song. FIGS. 11-13 includes a
screen display and flow charts associated with the style EQ function,
which displays the styles associated with the current playlist and allows
the subscriber to adjust the mix of songs played from the playlist.
The General User Interface
Before describing the preferred "more like" and "style EQ" functions in
detail, it will be helpful to understand the basic features of the
preferred audio on demand system and the user interface. This information
is provided in conjunction with FIGS. 4-6, which illustrate screen
displays that appear on the subscriber's display or monitor. Each of the
screen displays that form a part of the user interface provide information
and control objects, which typically appear as buttons. The subscriber may
select and activate the control objects using the directional control and
action button on the remote control unit (FIG. 3). An object is selected
by moving the "focus" to the object. The focus is analogous to a cursor on
a general purpose computer, and may be represented by a highlighted or
colored frame or border that appears around a control object.
The audio on demand system relies on an audio content database, which
includes all of the programming information items (e.g., songs) available
on the system. The songs in the audio content database are obtained from
various sources and are loaded on the continuous media servers that form a
part of the headend system. In most cases, the operator of the system will
arrange for record companies to provide their music catalogs. Those
skilled in the art will appreciate that while the audio on demand system
is capable of combining music catalogs from various sources into an
integrated music resource, some music companies may prefer that their
music not be mixed with music from other publishers. In this case, the
system may make different music catalogs available on different channels
on the interactive network.
FIGS. 4-6 illustrate the basic features of the audio on demand user
interface. FIG. 4 illustrates the initial audio on demand screen display
400, which appears when the subscriber selects the audio on demand system.
The initial screen display 400 is divided into three (3) different regions
or panels. The top region is a graphic display field 405, which may be
used to display graphic images associated with the audio on demand system.
For example, the service provider may choose to display a service mark,
trademark or other logo when an audio on demand channel is selected by the
subscriber.
Immediately below the graphic display field 405 is the style EQ panel 410.
The style EQ panel 410 is used to select the style equalizer, which is
described below. The style EQ panel 410 also includes a style EQ status
indicator 415, which indicates whether the style EQ feature is currently
activated.
The bottom region of the screen display 400 is used to select individual
songs or playlists. A find button 420 is provided in order to allow a
subscriber to select a specific song. When the find button is activated,
the screen displays bins that appear to be similar to those found in music
stores. In these bins, artists are listed in alphabetical order. When the
subscriber selects a particular artist, the names of the artist's albums
appear in chronological order, and are followed by an alphabetical listing
of the artist's songs. The subscriber may use this feature to select a
particular song or album. If the user selects a song, that song is loaded
into a new playlist. If the user selects an album, all of the songs from
that album are loaded into a new playlist.
The screen display 400 also includes a plurality of playlist buttons 425,
which allow the subscriber to select a playlist. As described briefly
above, a playlist is a collection of songs. Playlists may be generated in
a variety of different ways. For example, various types of playlists may
be provided by the service provider or other publishers. In addition, a
subscriber may build a playlist one song at a time using the find button
420. A subscriber may also communicate with the preferred interactive
network via a personal computer. When connected in this manner, the
subscriber may use the personal computer to create and name playlists,
perform abstract music searches or queries, etc. Those skilled in the art
will appreciate that the interface provided by a personal computer is much
more efficient for these tasks than a cumbersome on-screen interface that
relies on input from a remote control unit. Those skilled in the art will
also appreciate that playlists need not be limited to songs. Playlists can
include collections of news stories, movies, and other types of
programming information. The five playlist buttons 425 are similar to the
preset buttons on car radio and indicate the subscriber's five favorite
playlists.
The bottom portion of the screen display 400 includes an alphanumeric
display 430, which is used to provide some feedback when the subscriber
selects one of the other control objects on the screen. For example, when
the subscriber uses the remote control unit to move the focus to one of
the playlist buttons, the alphanumeric display 430 indicates the name of
the associated playlist. Likewise, when the find button has the focus, the
alphanumeric display 430 displays an informative message such as "select a
song".
FIGS. 5 and 6 illustrate the format of the screen displays 500 that are
associated with individual playlists. The playlist screen display 500
appears after the subscriber has created a new playlist by using the find
button 420 or selected an existing playlist using one of the playlist
buttons 425 on the initial screen display 400 (FIG. 4). Like the initial
screen display 400, the playlist screen display 500 includes a graphic
display field 505 and a style EQ panel 510. The graphic display panel 505
may be used to display general information associated with the playlist or
information associated with the currently playing album or song.
The bottom portion of the playlist screen display 500 includes a playlist
identification box 515. If the playlist was selected by choosing one of
the playlist buttons 425 on the initial screen display, the playlist
identification box 515 will include the same name, logo or icon that
appeared on the playlist button. In the preferred audio on demand system,
if the playlist was selected by using the find button 420 on the initial
screen display, the playlist identification box 515 will display an icon
that resembles a compact disc. The compact disc icon is used to indicate a
user preference playlist.
As mentioned above, a playlist is a collection of songs. When a playlist is
selected, the audio on demand system begins to play the first song in the
playlist. The name of the current song is displayed in a song title box
520. The artist's name is displayed in an artist box 525. A counter 530
displays the elapsed time of the current song.
The playlist screen display 500 also provides a list button 535, which may
be used to display a list of the songs that are included in the current
playlist and to jump to another song in the playlist. FIG. 6 shows a
screen display 600 with a pop-up list 605, which is displayed when the
subscriber activates the list button 535 on the playlist screen display
500. Each entry in the list includes the title of the song and the artist.
In the preferred audio on demand system, the list displays ten (10) of the
songs in the current playlist. The subscriber may use the directional
control on the remote control unit to scroll through all of the songs in
the playlist. The subscriber may also select any of the songs in the
playlist by using the directional control to highlight the desired song
and pressing the action button (on the remote control unit, FIG. 3). After
the subscriber selects a song from the list 605, the system returns to the
playlist screen display 500. At that point, the newly selected song begins
to play, and the song's title and artist are displayed in the song title
box 520 and artist box 525, respectively.
The playlist screen display of FIG. 5 also includes a "more" button 540, a
"like" button 545, and a "dislike" button 550. The "more" button 540 is
used to activate the "more like" music search function, which is described
below. The subscriber may add the currently playing song to a playlist
called "my favorites" by activating the "like" button 545 while a song is
playing. If the user does not like the current song, the subscriber may
activate the "dislike" button 550 while the song is playing. Once the
subscriber indicates the song is disliked, the audio on demand system will
never play that song again for the subscriber. This is true without regard
to where the song is found. In the preferred audio on demand system, the
only way for a subscriber to again listen to a song has been labeled as
disliked is to select that specific song using the find button on the
initial screen display (FIG. 4).
The "More Like" Function
Generally described, the "more like" function of the present invention
provides systems and methods for using a seed song (e.g., the current
song) to add new songs to a playlist. This is accomplished on the basis of
subjective style classifications and style weightings that are associated
with the songs in the audio content database.
The "more like" function allows for the context based selection of
subjective material. More particularly, the "more like" function allows a
subscriber to locate additional songs on the basis of subjective decisions
that have been made regarding the styles of the songs. In order to work
properly, the subscriber must be able to predict the output of the "more
like" function to some extent. In other words, the "more like" functions
must find songs that most subscribers would agree are "similar" to the
seed song.
Those skilled in the art will appreciate that it is subjective content that
complicates the classification of information. Systems that classify only
objective content are easily implemented. For example, songs are easily
classified and identified by their title and artist. However, systems that
accurately and predictably classify and search subjective content are more
complex. In this sense, the present invention is applicable to any systems
that classify and select programming information having subjective
content. However, in the preferred system, the invention is described in
the context of musical selections.
In the present invention, the subjective content associated with each song
is embodied in style tables, which are tools for classifying each song's
subjective content. Each song can be associated with any number of
different styles. The editor that creates the style table must determine
how important each style is to the description of each song. This is
reflected by weighting each style as it pertains to each song. Thus, the
process of creating a style table for an artist involves two steps: (1)
creating the list of possible style categories; and (2) assigning
weightings to each style category. Both of these steps are performed by
the editor that creates the style table.
The "more like" function relies on two elements: a database and a method
for searching and combining the songs in the audio content database. As
mentioned above, the database takes the form of style tables that classify
the style of each song. Although human beings work well as editors to
provide the required editorial content, there are limits as to the number
of categories that can be considered. For example, the operator of the
audio on demand system may deem it acceptable to spend the time to
classify the artists of the songs that are provided on the system.
However, the system operator may not consider it feasible to classify each
and every song on the system.
In the present system, these considerations are accommodated by allowing
music to be classified by various levels (e.g., artist, album, song). The
editor is responsible for choosing the particular level or levels that
will be used in the system. This decision depends on the amount of
editorial time the editor is willing to spend and the specificity required
for the desired outcome of the "more like" function. Although they require
more editorial work, higher level style tables (e.g., album or song) allow
the style tables to more accurately reflect the styles associated with
each song or album. This is especially useful in the case of artists whose
styles have varied over their career or from album to album.
In the preferred audio on demand system, the style tables are constructed
at the artist level, which is the lowest level, or broadest area of
categorization. This reduces the amount of editorial work required, while
providing style information that can be broadly applied to all of the
songs available on the system.
In the preferred system, the each song has a song identification (ID)
number that uniquely identifies that song. Similarly, each artist is
identified by a unique artist ID number. The digital audio data is stored
on a continuous media server by song ID number. The associated
administrative information is stored on an administrative server. The
administrative information includes the style tables, information for each
song (title, artist, album, etc.), and all of the other databases,
graphics, text, etc. that are required by the audio on demand system. A
playlist is created by creating a database that includes the song ID
numbers of the songs that are included in the playlist.
In the preferred embodiment of the present invention, the style tables
operate in the following manner. The audio on demand system operator
creates an artist level default style table for all of the artists whose
songs appear on the system. As mentioned above, the editor must determine
which style categories to use and the weightings assigned to each artist.
Therefore, the default style tables may include any number of style
categories associated with any number of artists.
An example of a style table for the Beatles is shown below:
______________________________________
Artist: The Beatles
Style Category Weight
______________________________________
1960s 1
1970s 1
British Invasion
7
Rock 5
Pop 5
Innovators 6
______________________________________
Although the audio on demand system provides default style tables for all
of the artists whose songs appear on the system, playlist publishers may
wish to provide their own style tables that categorize artists in a
different manner. For example, the default style tables may include a
single category for rap music. However, rap music aficionados may prefer
to further classify rap music into more precise subcategories, such as New
York City rap, Los Angeles Rap, Male Rap, Female Rap, etc.
The present invention allows playlist-specific style tables to be loaded
into the system with each playlist. Therefore, playlist publishers may
elect to use the default style tables, or may provide their own. Each
playlist-specific style table may reclassify all of the artists whose
music appears on the system, or only artists of particular interest. Thus,
in the previous rap music example, a publisher of a rap music playlist may
provide a style table that reclassifies those artists whose music appears
in the rap playlist. In other words, a playlist publisher can recategorize
the artists that are important to that publisher, and for which they want
to make finer distinctions.
The method of searching for and matching the entries in the audio content
database employs a qualitative scale of closeness, which is controlled by
a matching closeness indicator. In the preferred system, the matching
closeness indicator is a style slider, which is presented as part of the
user interface. The qualitative scale of closeness determines the degree
of similarity between the subjective content of the seed song and the
songs that will be chosen by the "more like" function. Thus, the style
slider allows the subscriber to determine the closeness of the match.
The advantage of the qualitative scale of closeness is that it purposely
allows the subscriber to choose a very broad setting. Those skilled in the
art will appreciate that many subscribers will want to expand their
playlists to include new songs that are only somewhat similar to earlier
entries. This provides a mechanism by which a subscriber may move his or
her playlist in new directions instead of repeatedly narrowing the
playlist. Likewise, this method of finding similar music is vastly
superior to searching by artist name or song title.
The outcome of the "more like" function depends on the relationship between
the number of styles in the style tables, the weighting scale, and the
position of the style slider when the "more like" function is activated.
FIGS. 7 and 8 illustrate the screen displays associated with the "more
like" music search function FIG. 7 illustrates the more like screen
display 700. A more like panel 705 appears when the subscriber activates
the more button 540 on the playlist screen display 500 (FIG. 5). The more
like panel 705 includes several control objects. A more album button 710
instructs the audio on demand system to list the other songs from the
album that includes the current song. A more artist button 715 instructs
the system to list more songs by the artist that performed the current
song. A yes button 730 and a no button 735 allow the subscriber to accept
or reject the list of songs that are presented as the output of the "more
album" and "more artist" functions. If the subscriber selects the yes
button, the listed songs are added to the current playlist.
A more style button 720 allows the subscriber to locate more music that is
like the current song. The more style button 720 operates in conjunction
with a style slider 725, which the subscriber sets to indicate the degree
of closeness or similarity that is required in order for a song to match
the current song. If the subscriber moves the style slider 725 all the way
to the right, the "more like" function will produce songs with styles that
are very similar to the current song. As the style slider 725 is moved to
the left, the "more like" function with present the subscriber with a
broader group of songs that are in some way similar to the current song. A
yes button 730 and a no button 735 allow the subscriber to accept or
reject the list of similar songs that are presented as the output of the
"more like" function.
FIG. 8 is a screen display 800 that includes a list 805 of songs that are
generated in response to the more style button. In the preferred system,
ten similar songs are presented to the subscriber. If the subscriber likes
the songs in the list, he or she can add the songs to current playlist by
activating the yes button 730. If the subscriber wants to see other songs
based on the same matching criteria, he or she may activate the more style
button while leaving the style slider in the same position. If the
subscriber wants to see songs that are more or less similar than those in
the current list, the subscriber may adjust the style slider and activate
the more style button. Finally, if the subscriber decides not to add any
songs to the current playlist, the subscriber may activate the no button
735. Once the subscriber activates the yes or no button, the list 805 and
more like panel disappear, and the system displays the playlist screen
display 500 (FIG. 5). The details of the "more like" engine are provided
below.
FIG. 9 is a flow diagram that summarizes the steps carried out by a
subscriber who is using the more button 540 to find more music. The method
900 begins at step 905 when the user selects a song to hear. This is
accomplished by using the find button to select a specific song, or by
selecting a playlist. Those skilled in the art will appreciate that the
"more like" function uses the current song as a "seed song" and selects
other songs that match the style criteria associated with the seed song.
At step 910 the subscriber activates the more button 540 on the playlist
screen display 500 (FIG. 5). This causes the more like panel to appear. At
step 915 the subscriber sets the style slider in order to determine the
closeness of the match. When the style slider is moved to the right, the
"more like" function finds songs whose style more strongly resembles the
seed song. As the style slider is moved to the left, the "more like"
function relaxes the degree of similarity that is required. The function
of the style slider is described more completely below.
At step 920 the user activates the more style button 720 on the more like
panel 705 (FIG. 7). When the subscriber activates the more style button,
the audio on demand system carries out the process of identifying songs
that have a style similar to the seed song. This process is described in
detail below in connection with FIG. 10. When the process is complete, the
system displays a list of ten (10) songs for review by the subscriber.
This list is reviewed by the subscriber at step 925.
At step 930 the subscriber determines whether the songs included in the
list should be added to the current playlist. If so, the subscriber
activates the yes button on the more like panel, and the system adds the
songs to the playlist (step 935). If the subscriber decides not to add the
songs to the playlist, the subscriber must decide whether to quit the more
like function (step 940). If so, the subscriber activates the no button
and the more like panel is dismissed (step 945).
At step 940 the subscriber may decide to try the more like function again
and see a different list of songs. If this is the case, the subscriber
determines whether to leave the style slider in the same place (step 950).
If so, the subscriber returns to step 920 and activates the more style
button. This causes the system to display ten other songs from the group
of songs that was generated earlier. If the subscriber decides to expand
or narrow the matching criteria, the subscriber returns to step 915 and
adjusts the style slider prior to activating the more style button.
FIG. 10 is a flow diagram illustrating the "more like" function of the
present invention as implemented in a program module running on a headend
server, which forms a part of the interactive network. The method 1000
begins at step 1005 after the user has selected a seed song and activated
the more style button. At step 1005 the system uses the style tables to
identify the style categories and weightings that are associated with the
seed song. In the preferred embodiment, which implements only artist level
style tables, this step involves identifying the style table that
corresponds to the artist that performed the seed song. The style table
data is then sorted by weighting in decreasing order.
At step 1010 the system determines the setting of the style slider. The
style slider operates to indicate a percentage, which is applied to the
sorted style table in the manner described below. The percentage
associated with a particular style slider position depends on the
granularity of the style slider. If the style slider has 11 positions, the
positions would represent increments of 10% each (ranging from 0% to
100%). Thus, the rightmost position would indicate a 100% match was
desired. If the style slider is in the center position, that would
indicate a 50% was desired.
At step 1015 the system uses the seed song's style table and the position
of the style slider to identify more songs that are like the seed song. In
the preferred system, this step involves identifying other artists who
have the same styles as the seed song artist at weights that are at least
as high as the position of the style slider.
The following example will illustrate the operation of this step, and the
relationship between the styles in the style table, the weighting scale,
and the position of the style slider. Assume the style table weighting
scale ranges from 1-10 and the style slider has 11 positions (at >0%, 10%,
20%, . . . , 80%, 90%, 100%). If the style slider is set at 100%, the
system will determine which of the style categories associated with the
seed song artist have weightings of 10 (which is 100% of the 1-10 scale).
After these style categories are identified, the system will search for
artists who have at least one of the same style categories at a weighting
of 10. Those skilled in the art will appreciate that this process will
yield artists whose style is very similar to the seed song artist.
If the style slider is set at 50%, the system will determine which of the
style categories associated with the seed song artist have weightings of
at least 5 (which is 50% of the 1-10 scale). After these style categories
are identified, the system will search for artists who have at least one
of the same style categories with a weighting of at least 5.
If the seed song is a Beatles song and the system uses the sample style
table provided above, a style slider position of 50% would match artists
who have the styles British Invasion, Innovators, Rock or Pop with a
weighting of at least 5. Those skilled in the art will understand that the
"more like" function is constrained by the style categories that are
associated with the seed song artist. However, the matching weightings for
those style categories are determined by the position of the style slider.
At this point, it should be appreciated that the style slider positions,
which are determined by the system provider, work with any style table,
regardless of the weighting scale used. For example, if a style table uses
a weighting scale of 1-100, a style slider position of 60% will search for
artists having the requisite styles with a weighting of at least 60. This
allows playlist publishers and others to create compatible style tables
using any size weighting scale.
After the matching artists are identified, the system compiles a list of
the songs performed by those artists. In order to limit the number of
songs that may be included in the group, the system can be designed to
select only a predetermined number of songs by each artist.
At step 1020 the system performs a random sort of the songs that were
identified in step 1015. At step 1025 the system picks the first ten songs
from the sorted group of songs and displays a list of those 10 songs to
the subscriber. This is illustrated in FIG. 8. In the preferred system,
the style categories and weightings that are used in the search are not
displayed to the subscriber.
At step 1030 the system determines whether the subscriber has accepted the
songs by activating the yes button on the more like panel. If so, the
method proceeds to step 1035, the 10 songs in the list are added to the
current playlist, and the method 1000 ends. Instead of adding all 10 songs
to the playlist, an alternative user interface may be provided in order to
allow the subscriber to specify which of the 10 listed songs should be
added to the playlist.
If the subscriber does not activate the yes button, the method proceeds to
step 1040 and determines if the subscriber has activated the no button on
the more like panel. If so, the method 1000 ends.
At step 1040 the system may determine that the subscriber has again
activated the more style button. If this occurs, the system proceeds to
step 1045 and determines the position of the style slider. If the position
of the style slider is the same as before, the system returns to step 1020
and resorts the same group of songs. If the subscriber moved the style
slider before reactivating the more style button, the system returns to
step 1015 and identifies other songs that match the new criteria.
Although the preferred system only implements artist level style tables,
the system could also implement album level style tables and song level
style tables. Those skilled in the art will appreciate that using "low
level" style tables (i.e., artist level) reduces the amount of editorial
work required to classify the music available in the system. While
requiring more editorial work, higher level style tables (e.g., album or
song) allow the style tables to more accurately reflect the styles
associated with songs or albums. This is advantageous because it can be
used to take into account artists whose styles have varied over their
career or from album to album.
If more than one level of style tables is provided, the system may be
designed to implement an "aggregation function" in the process of step
1015. Aggregation allows the system to combine one or more levels using
any type of mathematical operator. For example, "adding" style levels
leads to a tighter match between songs. "Multiplying" style levels results
in a broader spread of songs that will match the seed song.
Although the "more like" function has been described as searching the
entire audio content database, it is possible to limit the search material
that is searched by the system. For example, instead of searching all
published songs, the "more like" process may be used to search only new
releases. This would allow a subscriber to use the "more like" function to
add new music to a playlist. Those skilled in the art will understand that
the source material in the audio content database may be selected or
restricted in any number of ways, and that the data used to make such
distinctions is maintained on the interactive network's administrative
servers.
From the foregoing description of the "more like" function, those skilled
in the art will appreciate that the present invention includes two means
for providing context during the search. First, the style slider allows
the subscriber to control the closeness of the matches provided by the
"more like" function. Second, the present invention employs editorial data
produced by the system operator and playlist publishers to classify the
songs in the audio content database. Because new style tables may be
loaded in with a playlist, the outcome of the "more like" function will
vary depending on the nature of the style table and the editorial
decisions made by the playlist publisher.
Finally, those skilled in the art will appreciate that the present
invention provides distinct advantages over various other computer based
processes that could be used to identify similar songs. For example, it is
possible to implement a "more like" engine based on the computer analysis
of rhythm, tempo, etc. However, such an approach would require relatively
powerful computer processors, and would require that all of the songs in
the audio content database be pre-analyzed. Furthermore, such a system may
not be predictable, because most listeners would not equate jazz at 120
beats per minute with classical at 120 beats per minute.
The Style EQ Function
As described briefly above, the style EQ function addresses two distinct
needs that arise in the interactive network environment. These problems
arise because a subscriber typically selects a playlist on the basis of a
very short title, and because playlists may include a relatively large
number of songs. First, the style EQ allows the subscriber to get a
clearer look at what types of music are included in the playlist. The
system accomplishes this by displaying an indicator for each of the
predominant styles in the playlist and setting the position of the
indicators to reflect the relative portion of the playlist that includes
that style. This allows the subscriber to see how much music of each style
is present in the playlist. Second, the style EQ feature allows the
subscriber to alter the mix of the songs that are played from the playlist
by adjusting one or more of the indicators. Thus, if the subscriber does
not care for one of the styles in the playlist, the subscriber can
decrease the amount of that style that is played. Similarly, the
subscriber can boost the styles of music that he or she enjoys, which acts
as a filter and does not alter the actual content of the playlist. This
allows a subscriber to listen to a playlist in a variety of different
ways.
FIG. 11 illustrates the screen display associated with the style EQ
function provided by the preferred audio on demand system. The style EQ
screen display 1100 is displayed when the subscriber selects the style EQ
function from the style EQ panel on the initial screen display 400 (FIG.
4) or the playlist screen display 500 (FIG. 5). The style EQ screen
display 1100 covers the bottom portion of the display. The style EQ screen
display includes an on/off button 1105, an alphanumeric display 1110, and
a plurality of faders 1115. The style EQ is turned on and off by
activating the on/off button 1105. The alphanumeric display 1110 provides
information to the subscriber.
The preferred style EQ includes eight (8) indicators, or faders 1115. Those
skilled in the art will appreciate that the style EQ faders resemble a
conventional graphic equalizer. However, instead of each fader being
assigned to a frequency band, each fader is assigned to a particular style
of music included in the playlist. This allows the faders to be used to
give a subscriber a clearer picture of the types of music included in a
playlist. For example, a playlist that includes rock music may simply be
called "Rock". The style EQ faders may indicate that the playlist includes
music that may be more specifically described as 1970s rock, 1980s rock,
1990s rock, soft rock, acid rock, heavy metal, etc.
When a playlist is loaded and the style EQ function is first turned on, the
faders 1115 are positioned by the system to indicate the portion of the
playlist that fits into the associated style category. The subscriber may
get an idea of what is included in the playlist by using the remote
control unit's directional control to highlight each of the faders. The
display 1110 displays the name of the style associated with the
highlighted fader.
The style EQ function also allows the subscriber to adjust the mix of songs
that is played from the playlist. For example, if the subscriber dislikes
acid rock and heavy metal, the subscriber can "attenuate" those styles by
using the remote control unit to move those faders to their lowest
position. Likewise, the subscriber can "boost" the amount of soft rock
songs that are played by moving the fader upward. Those skilled in the art
will appreciate that the style EQ function does not alter the content of
the playlist. Instead, it merely adjusts the mix of songs that are played
from the playlist. The details regarding the operation of the style EQ
function and the assignment of style names to the faders are discussed
below.
FIG. 12 is a flow diagram that summarizes the steps carried out by a
subscriber who is using the style EQ function. The method 1200 begins at
step 1205 when the user selects a playlist. This is accomplished by using
the playlist buttons on the initial screen display. After a playlist is
selected the subscriber activates the style EQ screen display 900 (FIG. 9)
by activating the style EQ button on the playlist screen display (step
1210). This causes the system to display the style EQ panel with the
faders set to indicate the mix of songs included in the playlist.
At step 1215 the subscriber reviews the style labels associated with each
fader and the proportion of songs that are described by that style. At
step 1220 the, subscriber determines whether to adjust the sliders in
order to alter the mix of music that is played from the playlist. If so,
the subscriber uses the direction control on the remote control unit to
adjust one or more faders up or down (step 1225). The subscriber then
proceeds to step 1230 and dismisses the style EQ panel. If at step 1220,
the subscriber decides not to adjust the faders, the subscriber proceeds
to step 1230 and dismisses the style EQ panel.
FIG. 13 is a flow diagram illustrating the style EQ function of the present
invention as implemented in a program module running on a headend server,
which forms a part of the interactive network. The method 1300 begins at
step 1305 by playing a playlist that has been selected by the subscriber.
At step 1310 the system identifies the predominant styles of music that are
included in the playlist. Those skilled in the art will appreciate that
this step may be accomplished in a variety of ways. In the preferred
system, the style information used by the style EQ function is provided by
the publisher of the playlist, and is loaded into the system when the
playlist is selected. In this case, the playlist style data defines the
style categories that will be associated with each of the faders and
provides the initial settings for the faders.
In an alternative embodiment, the system may assign style categories to
faders by reading the style tables (provided in conjunction with the "more
like" function) and assigning the predominant style categories to the
faders. The style tables would provide the information necessary to
determine how many songs are associated with each style, and the relative
portions of the playlist that are described by each of these styles.
Once the predominant styles have been identified, the system proceeds to
step 1315 and assigns the styles to the faders. As described above, the
style EQ function in the preferred system includes 8 faders. Those skilled
in the art will appreciate that there are no inherent limitations on the
number of faders (and associated styles) that can be used in conjunction
with the style EQ function.
Those skilled in the art will appreciate that either method allows the
fader labels to be determined by the music in each playlist. This avoids
the problems that would arise if the system defined only a fixed number of
style labels that could be assigned regardless of the types of music in a
playlist. The present invention allows broad labels to be used for
playlist containing a broad mix of styles and specific labels to be used
for narrower playlists. For example, if a playlist included all of the
music in the world, the fader labels would be broad categories, such as
classical, jazz, country, rock, etc. Similarly, if a playlist includes
only jazz music, the style EQ function will assign meaningful jazz related
subcategories to the faders.
Defining the labels on the basis of the content of each playlist also
ensures that each fader label represents music that is in fact included in
the playlist. This avoids the problems of having a fader label without
having any music to go with it. For example, it would be misleading is
there is a standard label for jazz music, but a playlist does not include
jazz music. This would lead to the subscriber thinking he can increase the
amount of jazz music played from the playlist, when in fact the playlist
includes no jazz music.
At step 1320 the system adjusts the position of the faders to reflect the
relative portion of songs that are described by the style associated with
each fader. This allows the subscriber to see about what portion of the
playlist is represented by each style of music.
At step 1325 the system determines whether the subscriber has moved any of
the fader from their original positions. If not, the method proceeds to
step 1330 and plays all of the songs in the playlist in order.
If at step 1325 one or more of the faders have been moved, the system goes
to step 1335 and adjusts the mix of the music that is played from the
playlist. Those skilled in the art will appreciate that the style EQ
feature does not alter the playlist by adding or removing songs. Instead
it simply adjusts the mix of songs that are played from the playlist.
In the preferred system, the percentage of the songs that have each style
is determined by the following equation:
% of style=(value of style)/(total values for all styles)
In this equation, the value of each style is determined by the position of
the fader and the number of positions on each fader. For example, on the
style EQ panel of FIG. 10, each fader has 10 positions. If we refer to
each of the faders as styles 1-8 (from left to right), the total values
for all styles is 51 (which is the sum of 9 +3+7+6+4+6+7+9) out of a
possible 80. In this example, the percentage of music with style 1 is
9/51=18%. Similarly, the percentage of music with style 2 is 3/51=6%.
Those skilled in the art will appreciate that an advantage of the style EQ
feature is that moving a fader up or down leads to results that the user
can understand. Furthermore, although the style EQ function has been
described in the context of music playlist, those skilled in the art will
appreciate that this aspect of the invention may be applied to many
collections of material with subjective content. For example, the style EQ
could be applied to a playlist that includes news stories, with faders
labels such as national news, international news, business, sports, etc.
This would allow subscribers to adjust the faders so that they hear more
of the stories they are interested in, and less of the stories they are
not interested in.
From the foregoing description, it will be appreciated that the present
invention provides efficient systems and method for selecting and playing
music based on its subjective content.
The foregoing methods of the present invention may conveniently be
implemented in a program module that is based upon the flow charts in
FIGS. 10 and 13. No particular programming language has been indicated for
carrying out the various procedures described above because it is
considered that the operations, steps and procedures described above and
illustrated in the accompanying drawings are sufficiently disclosed to
permit one of ordinary skill in the art to practice the instant invention.
Moreover, there are many computers and operating systems which may be used
in practicing the instant invention and therefore no detailed computer
program could be provided which would be applicable to these many
different systems. Each user of a particular computer will be aware of the
language and tools which are most useful for that user's needs and
purposes.
The present invention has been described in relation to particular
embodiments which are intended in all respects to be illustrative rather
than restrictive. For example, although the present invention has been
described in the context of an interactive network system, those skilled
in the art will understand that the principles of the present invention
may be applied to, and embodied in, any type of interactive computing
device, including general purpose computers, personal computer, notebook
computers, etc.
Furthermore, the program modules described in conjunction with the
preferred embodiment run on the headend server, which forms a part of the
interactive system. Those skilled in the art will appreciate that the
system may be configured so that various program modules run on the set
top terminal. For example, data associated with the current playlist and
its style EQ settings could be downloaded to the set top terminal in order
to increase the processing speed.
Alternative embodiments will become apparent to those skilled in the art to
which the present invention pertains without departing from its spirit and
scope. Accordingly, the scope of the present invention is defined by the
appended claims rather than the foregoing description.
Top